-.
.
DOEAD113762
Stress and Permeability Heterogeneity within the Dixie Valley
Geothermal Reservoir: Recent Results from Well 82-5
Final Report - 05/06/1999 - 02/01/2000
S. H. Hickman
M.D. Zoback
C. A. Barton
R. Benoit
J. Svitek
R. Summers
December 1999
Work Performed Under Contract No. DE-FG07-991D13762
For
U.S. Department of Energy
Assistant Secretary for
Energy Efficiency and Renewable Energy
Washington, DC
By
Oxbow Geothermal Corporation
Reno, NV
DISCLAIMER
This report was prepared as an account of work sponsored
by an agency of the United States Government. Neither the
United States Government nor any agency thereof, nor any
of their employees, make any warranty, express or implied,
or assumes any legal Iiabiiity or responsibility for the
accuracy, completeness, or usefulness of any information,
apparatus, product, or process disclosed, or represents that
its use wouid not infringe privateiy owned rights. Reference
herein to any specific commercial product, process, or
service by trade name, trademark,
manufacturer,
or
otherwise does not necessarily constitute or impiy its
endorsement, recommendation,
or favoring by the United
States Government or any agency thereof. The views and
opinions of authors expressed herein do not necessarily
state or refiect those of the United States Government or
any agency thereof.
‘
—.
DISCLAIMER
Portions of this document may be iilegible
in electronic image products.
Images are
produced from the best available original
document.
.
DOEIID113762
STRESS AND PERMEABILITY HETEROGENEITY WITHIN THE DIXIE VALLEY
GEOTHERMAL RESERVOIR: RECENT RESULTS FROM WELL 82-5
FINAL REPORT
05/06/1999 - 02/01/2000
S. H. Hickman
M.D. Zoback
C. A. Barton
R. Benoit
J. Svitek
R. Summers
December 1999
Work Performed Under Contract No. DE-FG07-991D 13762
Prepared for the
U.S. Department of Energy
Assistant Secretary for
Energy Efficiency and Renewable Energy
Washington, DC
Prepared by
Oxbow Geothermal Corporation
Reno, NV
PROCEEDINGS. Twenty-Fifth Workshopon GeothermalReservoirEngineering
StanfordUniversity, Stanford,Cahfomia. January24-26.2000
SGP-TR- 165
STRESS AND PERNIEABILITY HETEROGENEITY WITHIN THE
DIXIE VALLEY GEOTHERMAL RESERVOIR RECENT RESULTS FROM WELL 82-5
S. H. Hickman’, M. D. Zoback2, C. A. Barton3,R. Benoit4,J. Svitekl and R. Summers’
‘U.S. Geological Survey, 345 Middlefield Road, Menlo ParlL CA 94025, U.S.A.
[email protected]
‘Department of Geophysics, Stanford University, Stanford, CA 94305, U.S.A.
[email protected]. edu
3GeoMechanics International, 250 Cambridge Avenue, Palo Alto, CA 94306, U.S.A.
[email protected]
40xbow Geothermal Corporation, 5250 South Virginia Street, Reno, NV 89502, U.S.A.
[email protected]
ABSTRACT
INTRODUC-TION
To understand causes for localized variations in stress
and permeability in a geothermal reservoir associated
with the Stillwater Fault Zone (SFZ), we collected
borehole televiewer, temperature and flowmeter logs
and conducted a hydraulic fracturing test in a well
(82-5) that penetrated the SFZ within the known
boundaries of the geothermal field but which failed to
encounter significant permeability. Although stuck
drill pipe prevented direct access to the SFZ, borehole breakouts and cooling cracks indicate a -90°
rotation in the azimuth of the least horizontal principal stress (S~~in)in well 82-5 at about 2.7 km depth;
similar rotations were observed in a producing well
located -0.6 km to the northeast. This rotation, together with the low Sm. magnitude measured at 2.5
km depth in well 82-5, is most readily explained
through the occurrence of one or more normal faulting earthquakes in the hanging wall of the Sl?Z in the
northern part of the reservoir. The orientation of Shmi”
below 2.7 km (i.e., -20 to 50 m above the top of the
SFZ) is such that both the overall SFZ and natural
fractures directly above the SFZ are optimally oriented for normal faulting failure. If these fracture and
stress orientations persist into the SFZ itself, then the
existence of a local stress relief zone (i.e., anomalously high Shnin magnitude) is the most likely
explanation for the very low fault zone permeability
encountered in well 82-5. Furthermore, under thk+
condition a massive hydraulic fracture within the SFZ
would propagate parallel to the strike of the SFZ and
possibly intersect the highly permeable res$rvoir to
the northeast. Thus, well 82-5 may be a good candidate for reservoir stimulation, although it would most
likely require that a new leg be drilled to regain
access to the SFZ.
Beginning in 1995, we have been conducting an
integrated study of stress and fracture permeability in
wells penetrating the Stillwater fault zone (SFZ) at
depths of 2 to 3 km (13gure 1). This fault is a major,
active, range-boundhtg normal fault that comprises
the main reservoir for a -62 MW geothermal electric
power plant at Dixie Valley, Nevada (Benoit, 1996).
Although earthquakes have not ruptured tlis segment
of the Stillwater fault in historic times, large (M= 6.8
to 7.7) earthquakes have occurred within the past 80
years along range bounding faults both to the northeast and southwest of the Dixie Valley Geothermal
Field (DVGF). Geologic evidence shows that the
Stillwater fault close to the DVGF experienced at
least two faulting episodes during the Holocene
(Wallace
and Whitney,
1984; Caskey and
Wesnousky, this volume).
The long-term goal of this study is to determine the
nature, distribution and hydraulic properties of fractures associated with the DVGF, and to chuacterize
the manner in which these fractures, and hence the
overall reservoir hydrology, are related to the local
stress field. This has involved conducting borehole
televiewer (BHTV) and temperature/pressure/spinner
(TPS) logging and hydraulic fracturing stress measurements in wells within the primary zone of
geothermal production (transmissivities -1 m2/min)
and in wells outside the boundaries of the DVGF that
were relatively impermeable (transmissivities -1 (IJ
m21min).These pre~ious results (summarized oeio w)
indicate that fault zone permeability is high only
when individual fractures as well as the overall
Stillwater fault zone are favorably oriented and critically stressed for frictional failure.
256
.. .,
%
.
,, ,
—
... .. ..... :’
. . . .. . . .
8
‘*
~’: r
*.,
~
: :.>., : .:.:
;:.
.: N .... .
4$
: .....
:, , :,:: “;$
:, j: I
~~‘“
.
..
~%i?
e“-’~]~cyw.
,&y;;;7:*7~::2K::;
;;:”
-
“’””i
●
m
eM;
-2?3
stress-induced borehole failure in BHTV logs from
wells within the permeable main reservoir indicate
that the local orientation of the least horizontal prin. cipal stress, S~~in, is nearly optimal for normal
faulting on the Stillwater fault. Although these
BHTV logs revealed pervasive macroscopic natural
fractures with a wide range of orientations, TPS logs
conducted during fluid injection and withdrawal indicate that the orientations of the most permeable fractures are distinct from the overall fracture population
and are subparallel to the Stillwater fault. Hydraulic
fracturing tests in these wells indicate that the magnitude of Sh~i. is low enough to lead to frictional
failure on both the main Stillwater fault and these
permeable fractures. Crack sealing and permeability
reduction would be expected along this segment of
the SFZ, given thermal and geochemical evidence for
up-dip transport of silica-saturated fluids. However,
the observation that highly permeable fractures
-within the DVGF are favorably aligned and critically
stressed for normal faulting suggests that dilatancy
due to intermittent fault slip is sufficient to counteract
the expected permeability reduction.
a7is~;Rwscy’v,
‘:”:;$$
.,, ..,:.,
+&y%
::: ;:&i6~’”
.: 4-?
:T
.... ..
&kjWA.&
; ...
Q1’”’
/
‘““’
... :. .:
i
“
%i
WTV,
‘“’”
f71X/EVALLEY
::<+
..5.
....-:
n
WELLS KEY
●
fRcwm14
9 Iwzcncw
e~ERVATl@4
i
4:1:
Figure
.,
:T~,
4
SHTV.
A SJiMLOW
WATER
——..
—
1. Map showing geothenna! wells tested in
Dixie Valley, Nevada. Injection and production wells penetrated
highly permeable
portions of the southeast-dipping Stillwater
fault zone, whereas observation wells failed
to encounter su.cient permeability to be of
commercial value. Measurements conducted
in these wells inc[uded TPS: Temperature/pressure/spinner logs; BHTV: borehole
televiewer logs; HFRAC: Hydraulic fracturing stress measurements.
Similar measurements were conducted in wells 66-21
and 45-14, which penetrated a relatively impermeable
segment of the SFZ approximately 8 and 20 km
southwest of the DVGF (Figure 1). The orientation of
S~~inin well 66-21 is near optimal for normal faulting
on the Stillwater faul~ but the magnitude of sh~i~is
too high to result in incipient frictional failure. In
contras% although the magnitude of S~~inin well 4514 is low enough to lead to frictional failure on optimally oriented faults, the Stillwater fault itself is locally rotated by 40° from the optimal orientation for
failure. This misorientation, coupled with an appar:nt
increase in the magnitude of the greatest horizontal
principal stress in going from the producing to nonproducing wells, acts to inhibit frictional failure on
the StiIlwater fault near well 45-14. Data from nonproductive wells located outside of the DVG,F thus
indicate that a necessary condition for high reservoir
permeability is that the SFZ be critically stressed for
frictional failure in the current stress field.
A number of “dry” legs have been drdled within the
known boundaries of the DVGF, where permeability
is expected to be high but, in reality, is low to nearly
nonexistent. In this paper, we present results from
BHTV logging and stress measurements conducted in
1999 in a well (82-5) located within the DVGF but
which failed to encounter sufficient permeability to
be economically viable (Figure 1). Well 82-5 was
drilled and then redrilled three times in 1985 and
1986 in an attempt to encounter
reservoir
permeabilities. The most recent (open-hole) leg
penetrated abundant sealed fractures starting at a
depth of about 2740 m - which we infer to be the top
of the SFZ - before passing through the main rangefront fault at 2833 m. Importantly, this well is located
only about 600 m southwest of some of the most
permeable production wells in the DVGF. The
primary goals of the well 82-5 study are to determine
why these pockets of low permeability exist within
the DVGF and if and in what manner massive
hydraulic fracturing or other reservoir stimulation
techniques might be used to turn 82-5 into a viable
geothermal production or injection well. TPS and/or
BHTV logs recently acquired from wells 25-5 and
27-33 (Figure 1), as well as TPS logs collected in
well 82-5, are still undergoing analysis and will not
be”discuised in detail here.
‘~THOD
The geometry and orientation of driiling-induced
tensile fractures, borehole breakouts and natural
fractures (see below) were determined using BHTV
logs. The BHTV is a wireline logging tool that provides a continuous, oriented, ultrasonic image of a
borehole wall (see Zemanek et al., 1970). The high
temperatures encountered in these wells (up to
-240° C) required the use of specialized televiewer
tools, owned by Stanford University and the U.S.
Geological Survey. These tools were repeatedly calibrated at the surface for orientation and ultrasonic
travel times in specially designed calibration tanks.
Data were digitized from the magnetic field tapes and
SUMMARY OF PREVIOUS RESULTS
The results of our prior investigations at Dixie Valley
are presented in Hickman et al. (1998), Barton et al.
(1998) and Morin et al. (1998). Observations of
257
.,
.
“.
.
—
.,
Di
DIXtE VALLEY WELL 82-6
Valley Well 82-5 Hyddrac at 2448 m
Fluid
2S9S
11
Injeaicm
l--f
~II h E--u.i?h
2S32
2596
II
2S33
2597
2S98
2S99
[k3+
.
1
0
Figure
2536
100
so
Elaped Tune, mbumas
150
2(
a)
2. Su@ace pressure and flow rate records
from the hydraulic fracturing test conducted
at 2448 m depth in well 82-5. Pressures
were also recorded using a downhole temperature/pressure/spinner
(TPS) tool suspended just above the kasing shoe at 2043 m
depth. This TPS tool was also used during
high-fiow-rate injection on cycles 4 and 6 to
identifi the exact location of the hydroj?ac.
26M
NESWN
Azlmu2h,
true
b)
NE
S*N
true
Azimuth,
Figure 3. Borehole televiewer log from well 82-5
showing: (a) borehole breakouts (irregular
dark patches) and (b) drilling-induced
tensile cracks (undulating vetiical features) on
diametrically opposed sides of the borehole.
Sev2ral natural fractures can also’ be seen in
(a) as sinusoidal alzrk lines (e.g., 2531.5 m).
tally determined utilizing a fracture initiation, or
breakdown, criterion for pure tensile fractures initiating in intact rock along the S-X direction. However,
as was the case in other wells tested at Dixie Valley,
BHTV logs conducted before the hydraulic fracturing
test showed that the open-hole interval in 82-5 contained numerous preexisting fractures (both natural
and drilling-induced) at a variety of orientations (see
below). Thus, it was not possible to directly measure
the magnitude of S~-. However, as discussed below,
bounds to the magnitude of SHm., were obtained using
estimates of compressive rock strength and the presence of borehole breakouts in well 82-5.
then processed to remove noise, stick-slip effects and
other tool-related problems.
Although hydraulic fracturing stress measurements
are typically conducted in short intervals of open hole
using inflatable rubber packers, high borehoie temperatures precluded the use of inflatable packers in
this study (see Hickman et al., 1988, for discussion of
the hydraulic fracturing technique and the interpretation methods used here). Instead, the hydraulic
fracturing test in well 82-5 was conducted by setting
a drill-pipe-deployed solid rubber packer in casing
and pressuring the bottom of the casing string and the
entire open-hole’ interw$’ to induce a hydraulic fracture in the uncased hole. Repeated pressurization
cycles were then employed to extend this fracture
away from the borehole. A TPS log. was conducted
over the open-hole interval of the well and into casing while pumping during the fourth and sixth cycles
of the hydraulic fracturing test to identify the depth at
which the hydraulic fracture was created (Figure 2).
The vertical (overburden) stress, Sv, was calculated
for well 82-5 using geophysical density logs conducted in nearby wells, in conjunction with rock
densities measured on surface samples obtained from
Dixie Valley (Okaya and Thompson, 1985).
RESULTS AND DISCUSSION
Unfortunately, during what should have been a routine work-over to cIear a blockage and condhion well
82-5 for testing, the drill pipe became inextricably
stuck and had to be severed at a depth of 2724 m,
preventing further access to the SFZ. Thus, although
BHTV and TPS logs and a hydraulic fracturing test
were obtained in the open-hoIe interval from 2071 to
2724 m, these data could not be used @ characterize
the state of stress, natural fracture population or permeability structure directly within the SFZ. Therefore, in the discussion that follows we of necessity
restrict our attention to observations made above the
SFZ and their implications for the nature of heterogeneities in stress and fault zone permeability within
The magnitude of .s~~inwas determined from the instantaneous shut-in pressure (ISIP), or the pressure at
which the pressure-time curve departs from an initial
linear pressure drop immediately after the pump is
turned off and the well is shut in (Figure 2). Pressures
recorded during a stepwise change in flow rate during
the last pumping cycle of the test were used to detect
changes in the permeability of the test interval resulting from closure of the hydraulic fracture and
provided an additional constraint on Sb~inmagnitude.
In a hydraulic fracturing test the magnitude of the
maximum honzontai principal stress, S-, is typi258
—
2000-
the DVGF and the potential for reservoir stimulation
in well 82-5.
. -SOuOm
Of casing
2100 -
;
Stress Orientations
Excellent quality BHTV logs were obtained over the
entire open-hole interval in well 82-5. They revealed
extensive stress-induced borehole breakouts and, to a
lesser extent, drilling-induced tensile (cooling) fractures (Figure 3). As discussed by Moos and Zoback
(1990), these tensile cracks result from the superposition of a circumferential thermal tensional stress
induced by circulation of relatively cold drilling
fluids and the concentration of ambient tectonic
stresses at the borehole wall. These tensile cracks –
which were also observed in nearby producing wells
73 B-7 and 74-7 (Hickman and .Zoback, 1997;
Hickman, 1998) - should form in a direction perpendictdar to the azimuth of S~i~. In contrast, borehole
breakouts form via compressive rock failure in response to tectonic stress concentrations at the
borehole wall and should be parallel to Shmi..
2200-
2300-
+
!
w
jij
.;
2400:
E
.
~ 2s00 4
!
:*
2700
2800
To=
i
2900
4
90
o
1
180
Breakout
Azimuth,
I
270
3
true
22007
Borehole breakouts were observed throughout much
of the logged interval in 82-5. They extend from the
top of the Miocene sediments at 2280 m to the top of
the stuck drill pipe at 2724 m, undergoing a sudden
-90° shift in orientation at about 2660 m (Figure 4a).
Tensile cracks were observed only between-2420 and
2620 m. The stress orientations indicated by the tensile cracks and breakouts in 82-5 are in very good
agreement (Figure 4b), and the azimuth of S~~i~above
2660 m is N23”E & 12° whereas below 2660 m it is
S66”E * 13°. Drilling-induced tensile cracks were
also observed in BHTV logs recently obtained in well
25-5, indicating that the azimuth of S~~inin that well
is S64”E & 14”.
2300-
2400-
m
b 2s00 *
z
x“
~ 2600E
+
2900
I
30
1
N
t
30
s ~~
I
60
Azimuth,
60
The azimuth of S~~i~above 2660 m in well 82-5 is
anomalous in that it is roughly parallel to the strike of
the SFZ, while the orientation of SR~inbelow 2660 m
(i.e., directly above the fault zone) is nearly perpendicular to the strike of the fault (Figure 5). The
deeper s~~i~direction from well 82-5 is thus in good
agreement with stress directions obtained in well
73B-7, in the producing interval of well 74-7
(Hickman et al., 1998), and in the nearby injection
well 25-5 (Figure 5). If this S~~i.direction observed
just above the SFZ persists to greater depths, then the
Stillwater fault where penetrated by well 82-5 would
be nearly at the optimal orientation for normal
faulting (Figure 4b).
30
true
Figure 4. (a) Orientations of borehole breakouts observed in the televiewer logs from well 82-5
(red dots). Calculated changes in breakout
orientation
with depth that would result
from a moderate-sized
normal ~aulting
earthquake on a fault passing through the
well are shown as multicolored bands (see
text). The depth extent of the damage zone
for the Stillwaterfatdt (sealed fractures] and
drill pipe stuck in the hole are also shown.
(b) Azimuth of the Least horizontal principal
stress, S~~i~,determined from breakouts and
drilling induced tensile crackr.
Stress Magnitudes
The well 82-5 hydraulic fracturing test shows that the
magnitude of S~~inis 31.1 ~ 0.6 MPa at a depth of
2448 m, corresponding to a ratio shim ls~ of 0.53
(Figure 6). Hydraulic fracturing tests from nearby
s!eothermal production wells 73B-7, 82A-7 and 37-33
~how that S~~,JSvranges from 0,45 to 0.51 at depths
259
.
.
..
?.
Stress, MPa
20 30
40
10
o
0
50
60
Oixie Valley Well 82-5
-1\\
‘~,
\\\\ k!2!2J
\\\\
\\
~KEY
500
70
I
1
Alluvium
1000
p.
E
=-
s,
\\
\\
\\
E 1500
&’
P -1’.0’.
..—
Uffs &
Tuft SUS.
\
2000%
\“:’\ \
3000 ! . . ...1....1..1,,.1,.,
‘A.swmmg C -80
>
Figure 5. Map of the Dixie Valley Geothermal Field,
showing the orientation of Shin as outwarddirected arrows. Also shown are lower
hemisphere
stereographic
projections
of
poles to permeable fractures
in selected
production wells, contoured using the Kamb
method (see Figure 8), demonstrating that
these fractures and the overall Stillwater
fauit zone are well oriented for normal
faulting in the current stress field (see
Barton et al., 1998; Hickman et al., 1998).
of 2.4 to 2.6 km, which is within a few hundred meters of the depths at which these wells penetrate
highly permeable fractures associated with the SFZ
(Hickman et al., 1998). Thus, although the S~~i~direction at the hydrofrac depth in 82-5 is highly
anomalous (Figure 4b), it is interesting to note that
the maximum differential stress (i.e., Sv - S~~i.) at
near-reservoir depths in this well is comparable to
that observed in nearby production wells.
Figure
,
As borehole breakouts were observed in the BHTV
log from well 82-5, a lower bound to the magnitude
of S~.,X was obtained using the S~~inmagnitude from
the hydraulic fracturing test together with theoretical
models for breakout formation. These models (see
Moos and Zoback, 1990) predict that borehole breakouts will initiate along the azimuth of S~~inwhenever
the maximum effective circumferential stress, (sWm,
at the borehole wall exceeds the compressive rock
strength, C,);i.e.:
6~-
= 3s~m - Shm,n
- Pp- Pm 2 c{,
‘“’’l’’”{
Ml% (MaceIw .Scdiments)
6. Magnitude of S~win,from the hydraulic
fracturing test in well 82-5. Also shown is a
lower bound on the magnitude of the greatest horizontal principai stress, S~mp based
upon the occurrence of stress-induced borehole breakouts in this well. The vertical
stress, Sv, and the formation jluid pressure,
PP, were calculated for the appropriate densities. The dashed lines indicate the range of
S~~i. at which incipient normal faulting
would be expected on optimally oriented
faiiltsfor coefficients offriction of O.6 -1.0.
cal well logs should be used to determine the appropriate value of COto use in Equation 1. Although we
hope to obtain more refined estimates for COin the
future, in the meantime we estimated this parameter
based upon published compilations of the compressive strength of rocks of similar lithology (Lockner,
1995). This analysis suggests that S~~.X is greater
than or equal to 49 MPa at a depth of 2.5 km (Figure
6). In contrast, no breakouts were observed in nearby
production well 73B-7, even though it exhibited
comparable S~~iOmagnitudes and penetrated the same
rock types as 82-5 (Hickman et al., 1998). This suggests that the magnitude of SH~aX in 82-5 is
significantly higher than in 73B-7, regardless of the
value of COused.
Using the Coulomb failure criterion (e.g., Jaeger and
Cook. 1976), one can estimate the critical magnitude
of S~~inat which frictional failure (normal faulting)
would be expected on optimally oriented faults using:
(1)
shmin‘~t
where PPis the formation pore pressure and Pmis the
mud pressure exerted on the borehole wall during
drilling. ideally, laboratory failure tests or geophysi-
=
(s” -
Pp)/ [(p~ + I)’n + Y]* + Pp
(2)
and assuming that the coefilcient of friction v ranges
260
—
BEFORE EARTHQUAKE
7“
Under these initial stress conditions, a normal faulting earthquake on either the SFZ or a subparallel
fault would result in a significant reduction in shear
stress on that fault (i.e., a coseismic stress drop).
Since Sv is fixed by the weight of the overburden,
this stress drop causes an increase in the magnitude
of Shtin. Aside from the small elastic (Poisson) coupling of changes in S~~,Xto S~~in,the magnitude of
SH~,Xremains relativelyunchangedsince it lies in the
Stress
\,\
,,,,,
\
1, ,
,,
,,
,,
‘,
; ‘\,%
‘=’ ‘? ;.6
~ip’,~
‘
AFfSR
s~
I
plane of the fault and contrihxes nothing to the shear
stress driving the earthquake. Given the nearly equal
\
magnitudes of the horizontal principal stresses prior
to the earthquake, the effect of the large coseismic
increase in &~lnrelative to that for SH_ is for s~i~ to ‘
transform into S~- and S~~mto WmSfOrnlintO &~in.
As can be seen through comparison of Figure 7 with
Figures 5 and 6, this model for the permutation of the
principal stresses can explain both the anomalous
S~~inorientation and the current magnitudes of Sti.
and S= above 2660 m in 82-5.
EARTHQUAKE
r
Stress
We used the.model of Barton and Zoback (1994) to
see if we could explain more quantitatively the stress
rotation seen in 82-5. This model superposes the
elastic stress field induced by slip on a fault of arbitrary orientation on the concentration of ambient
tectonic stresses around a borehole. The input stress
parameters for this model were chosen to be consistent with the pre-earthquake stress state depicted in
Figure 7, tied to the actual stress magnitudes
measured at 2448 m (Figure 6). The geometry of the
causative fault was prescribed to be parallel to the
SFZ, with the extent of the slipped patch (i.e., the
earthquake rupture dimensions) treated as free parameters and the coseismic stress drop assumed to be
total, The results of this model (Figure 4a) show
excellent agreement with-the orientation of breakoutk
observed both above and below 2660 m in well 82-5,
for a normal faulting earthquake extending from a
depth of about 2160 to 2690 m and piercing the borehole at 2380 m. Although not a unique solution, this
model also reproduces some of the important features
in the observed distribution of breakouts versus
depth. In particular, it predicts that breakouts should
be absent in the immediate vicinity of the stress rotation (i.e., from 2620 to 2700 m), where the
circumferential compressive stress should be near
zero (dark blue iine), and most strorigIy developed
adjacent to the midpoint of the rupture, where the
circumferential compressive stress should reach its
maximum value (dark red line).
Figure 7. Illustration of the manner in which a normal-faulting earthquake accompanied by a
large decrease in shear stress (stress drop)
on the causative fault could lead to a 90°
flip in the orientation of the two horizontal
principal stresses.
from 0.6 to 1.0, in accord with laboratory sliding experiments on a wide variety of rock types (Byerlee,
1978). As done for other nearby wells, we estimated
PP by assuming that it was in hydrostatic equilibrium
with the preproduction water table at 152 m depth
and integrating water density as a function of pressure and temperature, as appropriate to static temperatures measured in 82-5. This analysis indicates
that the Sb~inmagnitude at 2.5 km in 82-5 is low
enough to result in incipient frictional failure on optimally oriented normal faults (Figure 6). However,
as the orientation of Stii” at this depth is anomalous
and no hydraulic fracturing data are available from
greater depths in this well, it is not known if and to
what extent this condition of incipient frictional failure persists into the SFZ.
Model for Observed Stress Perturbations
A conceptual model for permutation (or exchange) of
the horizontal principal stresses due to a normal
faulting earthquake in the hanging wall of the SFZ
can explain the ah-rent stress state in well 82-5. In
this model (Figure 7), the state of stress before the
hypothesized earthquake was similar to that observed
today in well 73B-7 (Hickman and Zoback, 1997):
with Si.im perpendicular to the strike of the SFZ and
at the criticai magnitude for frictional failure. We
further propose, consistent with the upper bound to
S~~.X obtained for 73B-7. that SH~uXprior to this
earthquake was nearly equal in magnitude to S~~im.
Significant local stress rotations, perhaps due to slip
on nearby faults, were also observed in some of the
nearby production wells (Hickman et al.. 1998). Particularly dramatic stress rotations were seen in
production well 37-33, located about 0.6 km northeast of well 82-5 (Figure 5), where the direction of
S~min
changed by up to 90° over distances of a few ten
261
6$.
.
.
—
ure 3a). To facilitate comparison of these fractures
with the in-situ stress data, we have grouped them
according to whether they are above or below the
stress rotation at 2660 m (Figure -8). Above 2660 m
these fractures fall into two distinct populations: one
srnking between west and northwest with moderate
to steep dips (40° to 85”) to the south, the other is
more diffuse and srnkes between north and east with
moderate dips (45° to 75°) to the northwest. Although the west to northwest fracture set strikes
nearly perpendicular to the SFZ, it is near-optimally
oriented for normal faulting given the local azimuth
of S~~in(Figure 8a). Thus, fractures above 2660 m
appear to be dominated by normal faulting antithetic
to the Stillwater fault (i.e., at high angles to the SFZ).
This antithetic fracture set, which has not been observed before at Dixie Valley, presumably reflects a
mechanical overprint on the background fracture
population formed in response to the earthquakeinduced stress perturbation discussed above.
“Igure & Lower hemisphere, equal area projections
of poles to natural fractures observed in the
borehole televiewer log from well 82-5: (a)
above 2660 m and (b) below 2660 m. In the
Kamb contours to these poles (after Kamb,
1959), the density of shading is proportional
to the number of poles per unit area on a
lower hemisphere projection, nomudized to
the total fracture count, N. Also shown are
the pole representing the local orientation of
the Stillwater fault zone along with the azi- ,
muth of S%. over the corresponding depth
interval (see Figure 4bJ .
to a few hundred meters. A total of five stress measurements were conducted in well 37-33, indicating
relatively high S~~in/Sv values of 0.74 or greater at
depths of 1.57 to 1.89 km. These high S~~i~magnitudes are most likely due to moderate-sized earthquakes on faults within or adjacent to the SFZ, but in
this case S~~,X before the earthquakes would have
been cioser in magnitude to Sv than inferred for 32-5.
It is important to note, however, that the magnitude
and orientation of SW,ndirectly above the SFZ in well
37-33 (i.e., at a depth of 2.65 to 2.75 km) was such
that the SFZ and most of the associated permeable
fractures were critically stressed for frictional failure,
in spite of the dramatic perturbations in stress orientations and magnitudes observed in this well at
shallower depths.
Fracture Orientations
As observed in other nearby wells (Barton et al.,
1998), BHTV logs from well 82-5 show pervasive
macroscopic natural fractures with a wide range of
orientations throughout the logged interval (e.g., FIg-
The orientations of natural fractures below 2660 m in
well 82-5 (i.e., directly above the SFZ) are markedly
different from those observed above 2660 m. In particular, the west to northwest fracture set above 2660
m is notably absent in the deeper fracture population
(Figure 8b). Instead, these deeper fractures tend to
strike in a northeasterly direction and dip 40° to 75°
either to the southeast or the northwest, with the
dominant southeast-dipping set being subparallel to
the SFZ (green dot in Figure 8b). As the azimuth of
S~~i~below 2660 m is S66°E, this conjugate fracture
set is optimally oriented for normal faulting. The geometry of this conjugate fracture set is remarkably
similar to that observed for permeable fractures in
production wells to the northeast and southwest of
well 82-5 (Barton et al., 1998). As shown in Figure 5,
the dominant population of permeable fractures in
wells 74-7, 73 B-7 and 37-33 are parallel to the
Stillwater fault: striking northeast and dipping
roughly 50° southeas~ with a cqnjugate set striking in
roughly the same direction but dipping to the northwest.
Coulomb failure analysis (see Barton et al., 1998)
using the S~bn magnitude and orientation determined
in 82-5 above 2660 m shows that many of the natural
fractures in this interval are critically stressed for
frictional failure. However, as the magnitude of S~~in
below 2660 m is unknown, a corresponding analysis
could not be performed for the fractures shown in
Figure 8b. Thus, it is not known if and to what extent
the natural fracture population located immediately
above the SFZ is critically stressed for frictional
failure.
~
..
—
o
OixieValley Well 82-5: FluidLevelsw Tme
lea
200
Hyrkokac.
1
4
300 4
7W
800
1995
Figure 9. Illustration of the stress shadow model, in
which a chlorite-talc gouge with a low coefficient of friction (p) conned
to the main
shear zone of the Sliilwater fault might lead
to anomalously low shear stress on fractures
within the adjacent damage zone.
1996
1997
1998
1699
Year
Figure 10. Water level in well 82-5 before and ajler
the hydraulic fracturing test, compared to
the static reservoir level in nearby production well 45-33.
Potential for Reservoir Stimulation
Several factors suggest that well 82-5 may be a good
candidate for reservoir stimulation through massiV%-——
hydraulic fracturing. First of all, as noted above, it is
close to highly productive wells (e.g., 37-33) located
-0.6-0.9 km to the northeast. Secondly, it is in weak
pressure communication with, the reservoir. Third, the
bottomhole temperatures measured in the various
legs of 82-5 are similar to those measured in nearby
production wells at comparable depths. Thus, one of
the goals of this investigation was to evaluate the
feasibility of conducting a massive hydraulic fracture
in 82-5 to convert it into a viable production or injection well.
Indications for Low Fault Zone Permeability
The observation that well 82-5 is orders of magnitude
less permeable than the nearby production wells suggests that there may be fundamental changes in the
fracture population or the stress regime in proximity
to the SFZ over distances of less than 600 m that are
exerting a profound influence on well productivity.
As stuck drill pipe is covering the SFZ in well 82-5,
we cannot rule out anomalous fracture or stress orientations within the damage zone of the Stillwater
fault as the cause of low fault zone permeability at
this location based on our observations. However, we
consider this to be an unlikely explanation for this
low permeability ,because both the azimuth of Shni.
and the orientation of natural fractures directly above
the SFZ are nearly identical to those seen in nearby,
highly productive wells (cf., Figures 5 and 8b).
Since hydraulic fractures propagate in a plane perpendicular to the least principal stress, it is essential
that the azimuth of S~~inand the orientation and spatial distribution of permeable (or potentially permeable) fractures within the SFZ be accurately known in
evaluating the feasibility of reservoir stimulation
through massive hydraulic fracturing. Although we
were precluded horn making measurement directly
with the SF22 if the S~~iDazimuth observed dwectly
above the fault zone persists to greater depths, then a
massive hydraulic fracture within the SFZ would
propagate toward the highly permeable wells to the
northeast (Figure 5). Thus, well 82-5 may still be a
good candidate for massive hydraulic fracturing, although it would probably require that the well be
redrilled around the stuck drill pipe so that a massive
hydrofrac could be targeted directly within the SFZ.
In this regard, it is encouraging that the water level in
82-5 following our hydraulic fracturing test has declined below pre-test levels and is approaching that
recorded in nearby production well 45-33 (Figure
10). This suggests that our small-scale hydraulic
fracturing test enhanced the hydraulic connectivity of
well 82-5 to the overall geothermal reservoir.
As illustrated in Figure 9, we consider it more likely
that the low productivity of well 82-5 is due to localized increases in the magnitude of Swn (i.e., a reduction in shear stress) due to the presence of weak talc
within the main sheas zone of the Stillwater fault at
depth. In contrast to other (productive) wells drilled
within the DVGF, approximately 10 m of talc-rich
fault gouge was encountered on top of the granitic
footwall during drilling of 82-5. Laboratory strength
measurements show that talc is extremely weak, with
a coefficient of friction ranging from 0.2 to 0.25
(Morrow et al., 2000). Thus, the presence of weak
talc in the core of the SFZ at reservoir depths could
reduce the differential stress (Sv - S~~in)in the immediately adjacent country rock, effectively shielding
the potentially permeable (and now sealed) fractures
within the overlying damage zone from high tectonic
shear stresses. If this “stress shadow” hypothesis is
correct, then it suggests that the spatial extent of impermeable patches within the SFZ like that encountered by well 82-5 may be directly determined by the
distribution of talc (or other weak minerals) within
the main range-front fault.
263
“
.
$’.
.
CONCLUSIONS
REFERENCES
Analysis of borehole televiewer logs and a hydraulic
fracturing test in Dixie Valley weli 82-5 leads to the
foi~owing conclusions:
Barton, C., and M.D. Zoback (1994), “Stress perturbations associated with active faults penetrated by
boreholes: Possible evidence for near-complete stress
drop and a new technique for stress magnitude measurements”, Journal of Geophysical Research, 99,
9373-9390.
1. There is a -90° rotation in the azimuth of the least
horizontal principal stress, S~~i~$in this well at a
depth of 2.7 km. This rotation, which is similar to
stress rotations seen in producing well 37-33 lbcated
about 0.6 km to the northeast, is best explained
through a permutation of the horizontal principal
stresses in response to a moderate-sized normal
faulting earthquake on a fault subparallel to the
Stillwater fault zone (SFZ). These stress perturbations suggest that caution be used in extrapolating
shallow stress measurements to reservoir depths in
seismically active areas.
Barton; C.A., S. Hickman, R. Morin, M.D. Zoback
and R. Benoit (1998), “Reservoir-scale fracture permeability in the Dixie Valley, Nevada, geothermal
field”, Proceedings 23m Workshop on Geothermal
Reservoir Engineering, Stanford Univ., Stanford, CA,
299-306.
Benoi~ W.R. (1996), “Injection of geothermal fluid
in Nevada as typified by the Dixie Valley project”, in
~ep Injection Disposal of Hazardous and Industrial
Wastes, J. Apps and C.-F. Tsang (eds.), Academic
Press, San Diego, pp. 449-464.
2. The magnitude of S~~i, at 2.5 km depth is about
0.53 of the vertical stress, in accord with Byerlee’s
law for normal faulting. Since stress-induced borehole breakouts were observed in this well, this
suggests that the horizontal differential stress prior to
the hypothesized earthquake was quite low (as previously proposed for nearby well 73B-7, located about
2.5 km to the southwest) and that this earthquake was
accompanied by near-total stress drop.
Byerlee, J.~. (1978), “Friction of rocks”, Pure and
Applied Geophysics, 116,615-629.
- Hickman, S., and M.D. Zoback (1997), “In-situ stress
in a fault-hosted geothermal reservoir at Dixie
Valley, Nevada”, Proceedings 22’~ Workshop on
Geothermal Reservoir Engineering, Stanford Univ.,
Stanford, CA, pp. 141-146.
3. Although stuck drill pipe prevented us from making measurements directly within the SFZ, fracture
and stress orientations immediately above the fault
zone are similar to those seen in highly permeable
wells (73B-7 and 74-7) located 2-3 km to the southwest. In particular, both the SFZ and the overlying
natural fractures are optimally oriented for normal
faulting. Thus, a local stress relief zone (i.e., an
anomalously high S~~inmagnitude) within the SFZ is
probably the most likely explanation for the very low
permeability encountered in well 82-5.
Hickman, S., M.D. Zoback and R. Benoit (1998),
“Tectonic controls on reservoir permeability in the
Dixie Valley, Nevada, geothermal field”, Proceedings 23~ Workshop on Geothermal Reservoir Engineering, Stanford Univ., Stanford, CA, pp. 291-298.
Hickman, S., M.D. Zoback and J.H. Healy (1988),
“Continuation of a deep borehole stress measurement
profile near the San Andreas Fault, I: Hydraulic
fracturing stress measurements at Hi Vista, Mojave
Desert, CA”, Journal of Geophysical Research, 93,
15183-15195.
4. If the Sti, orientation observed in the bottom of
this well persists to greater depths, then a massive
hydraulic fracture within the SFZ would propagate
toward highly permeable wells to the northeast. Thus,
82-5 may still be a good candidate for reservoir
stimulation, although it would probably reipire that a
new leg be drilled to regain access to the SFZ.
Jaeger, J.C., and N.G.W. Cook (1976), Fundamentals
of Rock Mechanics, 2nd cd., 585 pp., Chapman and
Hall, London.
Kamb, W. B. (1959), “Ice petrofabric observations
from Blue Glacier, Washington, in relation to theory
and experiment”, Journal of Geophysical Research,
64,1891-1910.
ACKNOWLEDGMENTS
This work was supported by the U.S. Department of
Energy (DOE) Geothermal Technologies
and
Enhanced Geothermal Systems Programs under
Grant Number DE-FG07-991D 13762. However, any
opinions, findings, conclusions, or recommendations
expressed herein are those of the authors and do not
necessarily reflect the views of the DOE. Additional
support was provided by the Earthquake Hazards
Program of the U.S. Geological Survey.
Lockner, D.A. (1995), “Rock
Physics and Phase Relations: A
cal Constants,
T. Ahrens
Geophysical Union, Washington,
Failure”, in Rock
Handbook of Physi(cd.), American
DC, pp. 127-147.
Moos, D., and M.D. Zoback (1990), “Utilization of
observations of well bore failure to constrain the orientation and magnitude of crustal stresses:
264
-.
. ●..’
‘e—
Application to continental, Deep Sea Drilling Reject,
and Ocean Drilling Program boreholes”, Journal of
Geophysical Research, 95,9305-9325.
Morin, R., S. Hickman, C. Barton, A. Shapiro, R.
Benoit and J. Sass (1998), “Hydrologic properties of
the Dixie Valley, Nevad& geothermal reservoir from
well-test analyses”, Proceedings 23d Workshop on
Geothermal Reservoir Engineering, Stanford Univ.,
Stanford, CA, 307-314.
Morrow, C.A., D.E. Moore and D. A. Lockner
(2000), “The effect of mineral bond strength and adsorbed water on fault gouge frictional strength”,
Geophysical Research Letters, 12 pp. (in press).
Okaya, D.A., and G. Thompson (1985), “Geometry
of Cenozoic extensional faulting: Dixie Valley,
Nevada”, Tectonics, 4,107-125.
Wallace, R.E., and R.A. Wldtney (1984), “Late
Quatemary history of the Stillwater seismic gap,
Nevada”, Bulletin of the Seismological Society of
Americ% 74,301-314.
Zemanek, J., E. E. Glenn, L. J. Norton, and R. L.
Caldwell, (1970), “Formation evaluation by inspection with the borehole televiewer”, Geophysics, 35,
254-269.
265